chore: sync public mirror from internal - #1346
Conversation
|
Resolved 0 review-bot thread(s) on this generated public mirror PR. Review of mirrored code belongs on the mono source PR. |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: aacf8649a6
ℹ️ About Codex in GitHub
Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".
| context_window: Some(args.context_window), | ||
| cwd: cwd.path().to_string_lossy().into_owned(), | ||
| system_prompt: Some("Answer from the supplied conversation evidence. Return only the JSON object requested by the final question.".into()), | ||
| thinking_enabled: false, |
There was a problem hiding this comment.
With thinking_enabled disabled, NativeAgentRunner::build_config assigns primary requests temperature: Some(0.7), while this evaluator runs each arm only once and exposes no temperature/seed control. For provider/model routes that honor temperature, the original and compacted probes are therefore independent samples, so a nonzero difference_percentage_points can be caused solely by sampling variance rather than compaction. Set deterministic primary decoding (or record/configure repeated seeded trials) before treating the paired report as a compaction comparison.
Useful? React with 👍 / 👎.
Summary
dx-corp/monodx-corp/codeas a generated public mirror of the private source of truth38940ee14ede4a1d8f99729c6991b805720314b266915470e59f656745b3a6cfd9ab30f6db2273e512file(s) to copy/update and0stale file(s) to delete0sync-holdlabel to this PR; the sync workflow skips every push while it is setSource-of-truth status
Projection: deixic-code
Source: dx-corp/mono@38940ee14ede4a1d8f99729c6991b805720314b2
Prior destination base: 6691547
Content SHA-256: f199ddc735eb038778000c80291490a41682e110e834b8c98f37ca52cc0f3744
Result: drift_detected; 12 changed, 0 deleted
Destination-owned content is preserved. Destination CI is a separate health signal.
copy/update .repository-projection.json
copy/update evals/compaction-behavior-v1.json
copy/update packages/local-host-rs/Cargo.toml
copy/update packages/local-host-rs/examples/compaction_eval/main.rs
copy/update packages/local-host-rs/examples/compaction_eval/provider_tests.rs
copy/update packages/local-host-rs/examples/compaction_eval/report.rs
copy/update packages/local-host-rs/examples/compaction_eval/suite.rs
copy/update packages/local-host-rs/examples/compaction_eval/tests.rs
copy/update packages/local-host-rs/examples/compaction_eval/trial.rs
copy/update packages/local-host-rs/src/agent/mod.rs
copy/update packages/runtime-rs/src/agent/native/context.rs
copy/update packages/runtime-rs/src/agent/selective_summary.rs
Drift sample
Public-only commits since last generated sync
Validation
Test Plan
maestro-sync-public-release-mirrorworkflow inpublic-treemoderequire-internal-prconfirms internal source PR lineageStaged Rollout
dx-corp/mono@38940ee14ede4a1d8f99729c6991b805720314b2, including existing hidden/evaluation surfaces, and keeps public package parity behind the established public-source-provenance gate.